How Does Face Recognition Work?
Facial recognition technology is more than a simple face scanner or face match program.Facial recognition systems use a number of measurements and technologies to scan faces, including thermal imaging, 3D face mapping, cataloging unique features (also called landmarks), analyzing geometric proportions of facial features, mapping distance between key facial features, and skin surface texture analysis.
Facial recognition software is used in a variety of ways, but most often for security and law enforcement purposes. Airports use facial recognition software in a couple of different ways, such as scanning faces of travelers to search for individuals suspected of a crime or on a terrorist watch list and also to compare passport photos with in-person faces to confirm identity.
Law enforcement uses facial recognition software to identify and apprehend people who commit crimes. Several states use facial recognition software to prevent people from getting fake identification cards or driver’s licenses.
Some foreign governments have even used facial recognition technology to crack down on voter fraud.
Limitations of Facial Recognition
While face recognition programs can use a variety of measurements and types of scans to detect and identify faces, there are limitations.Poor resolution images and poor lighting can reduce the accuracy of face-scanning results.
Different angles and facial expressions, even a simple smile, can pose challenges for face matching systems.
Facial recognition loses accuracy when the person is wearing items like glasses, hats, scarves, or hair styles that cover part of the face. Makeup and facial hair can also pose issues for face detection programs.
Facial scans don’t necessarily connect with a profile, meaning that a scan of a person’s face may not be useful if there are no photos of them in an accessible database. Without a match, the identity of the person behind the face scan can remain a mystery.
Concerns over privacy or security can also pose limitations for how facial recognition systems can be used. For example, scanning or collecting facial recognition data without a person’s knowledge and consent violates the Biometric Information Privacy Act of 2008.
Also, while lack of a facial recognition match can be useless, a strong one can be a security risk. Facial recognition data that positively matches online photos or social media accounts could allow identity thieves to gather enough information to steal a person’s identity.
Facial Recognition Use in Smart Devices and Apps
Facial recognition is a growing part of our everyday lives through devices and applications.For example, the Facebook facial recognition system, DeepFace, can identify human faces in digital pictures with up to a 97 percent accuracy rate. And Apple has added a facial recognition feature called Face ID to the iPhone X. Face ID is expected to replace Apple’s fingerprint scanning feature, Touch ID, giving users the option of face login to unlock and use their iPhone X.
As the first smartphone with a built-in facial recognition feature, Apple’s iPhone X with Face ID is a good example to explore how facial recognition can work on our everyday devices. Face ID uses depth perception and infrared sensors to ensure the camera is scanning your actual face and not a photo or 3D model.
The system also requires your eyes to be open, to prevent another person unlocking and accessing your phone if you are asleep or unconscious.
Face ID also stores a mathematical representation of your face scan in a secure location on the device itself to prevent someone from accessing a photo of your facial recognition scan and prevents potential data breaches that would release this data to hackers because it doesn’t get copied to or stored on Apple’s servers.
Though Apple has provided some information on limitations of the Face ID feature. Children under 13 are not good candidates to use this technology because their faces are still growing and changing shape. They’ve also cautioned that identical siblings (twins, triplets) would be able to unlock each other’s phones. Even without an identical sibling, Apple has estimated that there is approximately a one in a million chance that the face of a complete stranger will have the same mathematical representation of their facial scan as you do.
